▲ 42 points
back
31 comments
I've been on sales calls and in PoCs with a couple major Mexican conglomerates (think Cemex, Pemex, or Azteca level or the MX operations for BBVA or Santander) so I've had the chance to interact with a number of Mexican software engineers, plus I have a couple Mexican diaspora SWE friends and honestly the talent is there, but it won't really be the next "Poland" or "India" like how the article portraies it.
The issue I've seen is you're going to be paying Austin in the 2010s level salaries ($60-70k base plus stock) for competitive talent due to remote work or Central Europe salaries at the lower end ($30-40k base), plus the added salary pressure of MX nationals being able to move to the Bay Area, Austin, NYC, or Seattle via the TN visa (most undocumented immigration coming from MX are not MX nationals - they're primarily Honduran, El Salvadoran, Nicaraguan, or Guatemalan nationals).
With that level of salaries, you aren't getting a significant discount compared to other locations nor the ability to run 24 hour operations. American companies are fine paying MX level salaries if not higher when you factor in equity in India, Romania, and Poland or pay US level salaries in Israel because it gives them the ability to have continuous 24/7 operations and a level of redundancy from a risk planning standpoint.
None of this is to say MX ain't a bad spot to develop talent - anything but given the free trade agreements with the US, South America, and the EU and the very American-esque regulatory system making it a great hub for American companies to expand within the Hispanoblante world - but the scene is similar to Germany or France's in the sense that the dev industry seems heavily Finance, Manufacturing, Defense, or Conglomerate driven and treats it as a cost center.
What's the benefit of 24-hour operations in software engineering?
I once worked for a US-based software company that opened an engineering office in Western Europe, planning to have engineers hand off features from one office to the next and keep development going around the clock. Of course this didn't happen because of the overhead of the daily handoff and the aggravation of your code never being in the same state in the morning that it was in the night before.
What actually happened was that both offices worked independently on different features, which was a perfectly productive arrangement, but there was no particular benefit to the fact that they were doing it "around the clock".
I always suspected this plan was just an excuse to seek cheaper talent outside the U.S. without offending the American developers, but now that I've seen someone else mention it, I wonder if there's a real benefit that I'm missing.
24 hour operations matter in large budget (>$20k/year) deals. There are SLAs to hit, massive penalties if they are not hit, and customers are VERY trigger happy with filing Severity 1 escalations and texting your company's C-Suite, Board, and Sales Leadership to get a bug resolved ASAP.
In additon, some level of regional redundancy is critical to prepare for black swan events (eg. a regional power outage like in the NE in 1996, flooding in Sacramento or East Bay suburbs leading to Product Owners or feature owners dealing with evacuations - this happened to us with the current flooding).
And finally, 24 hour operations allows you to expand to markets outside the US. While engineers in San Jose are sleeping, it's the middle of the workday for customers in London, Paris, Istanbul for example and they will inevitabely have bugs and issues arise.
Also, like I mentioned in the comment below:
2. Training and mentoring is a reflection of your company's Engineering Management. My employer has had an easy time offshoring development to Slovakia and Bangalore because we had Tech Leads and Engineering Management make the initiative to mentor, train, and onboard offshore team employees as we would domestic employees. I've seen similar examples occur across the industry (eg. Google Pay's dev team in Hyderabad, DE Shaw's dev team in Gurgaon, Motive/KeepTruckin dev teams in Lahore & Islamabad). If Management views Engineering as a cost center (which, it pains for me to say as an ex-cybersecurity engineer, at times it can be), that's where you tend to see the relationship between offshore and onshore teams collapse, because the relationship is based on the implicit arbitrage of labor.
Fair points. Continuous operations create significant drag in day-to-day feature development work, but I can see the value for truly urgent things like Sev-1s, or having engineers that can jump on customer calls in different timezones.
Definitely agree on the training point. That was one of the things we did right. I and several of the other senior engineers rotated through the foreign office for the better part of a year to get the new teams spun up on our codebase. Both offices maintained a good relationship afterwards.
Thanks for the clarification.
For sure. As an early stage startups (pre-Series D) that target lower market or mid-market (B2C startups like Chime or B2B SaaS startups like many of the ones you see in YC), continuous operations aren't significantly important at their stage, but as their revenue grows, it becomes more and more critical.
1. I don't think there would be an intention to move folks from MX to the bay area, Austin, etc. so not sure why you mentioned that.
2. I think having folks "near shore" makes it far easier to train, mentor, manage, etc. when there is a lot of overlap in time zones. When I had a team in India that was THE major challenge. The speed of iteration was abysmal.
3. I may be wrong but I have some nephews by marriage getting STEM degrees in Tijuana (TJ) and salaries for a new mechanical engineer hire are terrible IIUC. Per https://www.salary.com/research/mx-salary/benchmark/mechanic... average annual salary is 324,000 MX Pesos which is < $17000 USD. Software is similar: https://www.salary.com/research/mx-salary/benchmark/mechanic... . And TJ has a higher cost of living than other cities in MX and thus salaries due to proximity to the US. It will take years for supply to meet demand but I can see Mexico and South America being good sources of talent in the future. To me it's just a marketing, education and pipeline problem.
> 1. I don't think there would be an intention to move folks from MX to the bay area, Austin, etc. so not sure why you mentioned that.
It's actually quite easy to do it. For real engineers at least.
1. It isn't companies moving employees - it's Mexican nationals applying to jobs in the US and then moving. So long as you have working English fluency and a relevant degree, it's pretty trivial to get a TN, H1-B, or EB-1 as a Mexican national.
2. Training and mentoring is a reflection of your company's Engineering Management. My employer has had an easy time offshoring development to Slovakia and Bangalore because we had Tech Leads and Engineering Management make the initiative to mentor, train, and onboard offshore team employees as we would domestic employees. I've seen similar examples occur across the industry (eg. Google Pay's dev team in Hyderabad, DE Shaw's dev team in Gurgaon, Motive/KeepTruckin dev teams in Lahore & Islamabad). If Management views Engineering as a cost center (which, it pains for me to say as an ex-cybersecurity engineer, at times it can be), that's where you tend to see the relationship between offshore and onshore teams collapse, because the relationship is based on the implicit arbitrage of labor.
3. Tijuana is not a software hub. It is a medical device manufacturing hub though (all the design and R&D is done across the border in San Diego and manufacturing is done in Baja California). Also the MechE salary in MX is apparently higher than it is in Poland (~$13k/yr - https://www.salary.com/research/pl-salary/benchmark/mechanic...), Romania (~$4k/yr - https://www.payscale.com/research/RO/Job=Mechanical_Engineer...), or China (~$15k - https://www.salary.com/research/cn-salary/benchmark/mechanic...).
In fact, I think the Poland salary is overestimated as I know for a fact starting salaries for new grad SWEs in Poland (not working at MNCs) is around $7-8k/yr. Most of the software roles in MX are in CDMX, Monterrey, or Guadalajara. So in fact devs in MX might be earning more than in Eastern Europe.
I know for a fact McKinsey pays Associates 33% more in Mexico ($127k) than they do in Poland ($96k) for the exact same role, and generally McKinsey salaries take regional job markets into account when setting salaries, so I guess it shouldn't surprise me that Dev salaries might be higher in MX than PL or RO (and they 100% are higher than in UA or RU)
> MX nationals being able to move to the Bay Area, Austin, NYC, or Seattle via the TN visa
TN visa requires candidates have the "correct" university education (self-taught and music majors need not apply). Aside from that, you're absolutely correct.
I'm sure that the government is keeping track of how many people are getting TN visas per year, but otherwise these people just blend in and become part of the normal workforce. Eventually many of them get a green card or US citizenship.
They def can, though one of my diaspora buddies is seriously considering returning to CDMX even though he's making good money in SV, graduated from a top CS program in the US, and already got his GC.
To him the QoL in CDMX was better than other cities in the US (but homesickness is also playing a part)
Completely anecdotal, but about 8 years ago we had a pretty well scoped out web dev project. We didn't want to go with FTEs because the roadmap for that feature wasn't obvious after initial release (something bespoke for a large customer).
Someone at the company referred his "friends and college acquaintances" who ran a small consulting firm in Mexico. The project lasted around 6 months, they delivered a little ahead of schedule and were quite nice to work with. A team of 5 devs.
Looking at their LinkedIn profiles, two of them are now in the Bay Area for sure, one I'm not completely sure because I think his wife and him lived in California for a while since she did a postdoc there but I don't know if they stayed. One is in Seattle and the last is in NYC.
Every time someone tells me they found "cheap, local talent" in LATAM for "much less than American developers" I always wonder how long it will be until that talent becomes itself... an American developer! Especially since, unlike a decade ago, people are now incorporating in LATAM and hiring FTEs, not contractors. This makes them eligible to get an L-1 visa to transfer to the parent company in the United-States after ~1 year at the company.
I am from Poland and to be honest I always wondered why US startup outsource their "R&D" to central europe instead of latin america. The time zone difference just makes this unpleasant to both sides. I come to the conclusion that it must be the EU membership that provides some guarantees around intellectual property and private property rights, while at the same time the EU countries are safe to stay and offer at least a second world life for higher ups that need to bootstrap the local subsidiary.
In India we have a lot of cheap labor, but the air pollution (Bangalore anyone?) and often poor service quality (hotels? credit card scams?) makes it difficult for managers to feel at home.
I haven't been in Latin America yet so I cannot say how it compares to poorer EU countries but I can see that it has a lot of potential. Probably the biggest problems are political instability (populism; money printism and occasional bankruptcy) and crime. I think that Latin America could create something similar to EU, it would even more suite them as the language barrier would not be that big (2 major languages only). With shared laws and maybe a currency they could become a significant player on the world stage, but I am not sure if they have politicians that would pursue that vision.
Anyway, I wasn't there so it is only a prediction from antipodes...
> Rest of World spoke to 18 entrepreneurs, recruiters, and developers in Mexico, Peru, and Uruguay, all of whom acknowledged that the recent layoffs across the tech sector globally haven’t freed up experienced developers for hire by regional companies.
Tech companies don't hire and layoff exclusively tech workers. It takes more than engineers to run and grow a company. That is, when we hear X is laying off 15%...that's 15% of all bodies (e.g., marketing, HR, etc.) not all 15% is engineering.
Once you start dismissing engineers, you suffer a brain drain. I've seen nothing to indicate that's happening in general). Of course there might be a couple with no choice.
I do find it funny that the article basically arrives at "it's hard to find people to do this stuff" when the headline would imply the opposite.
At my last job we switched from contractors in India to a group in Peru. I wasn't around long enough to determine if there was a quality difference but being in the same time zone improved communications dramatically. PMs in India were constantly trying to reverse the relationship and have us work in the middle of the night if there was a need for a conference call. Our Peruvian colleagues started their days at the same time we did, so there was never an issue over which group should lose sleep for a conference call.
Take it or leave it. US got too expensive to produce and cannot compete with a growing number of correctly prepared engineers willing to work hard and cheaper abroad.
To close frontiers and creating new taxes won't be a solution this time. It's just a matter of time for a huge crisis in the industry.
While agreeing on the cost of living, I have to say that the number of engineers will not be growing too much (I say this as a latin american engineer): Most latin american countries are below the replacement level fertility of 2.1, especially the ones who provide the best quality engineers (Brazil, Argentina, Chile and Uruguay) and the ones who aren't such as Bolivia or Ecuador, are barely above replacement. Already college cohorts are 25% smaller than their historic maximum in Chile, and Chile together with Uruguay are facing a dramatic threat of decline since their total fertility rates are lower than 1.3 for 2022 and falling (for Argentina and Brazil it's no more than 1.65 for the same year, just below the TFR of the United States at 1.66).
As a result of all this, the economic structure of these aging countries is changing and the decreasing cohorts of engineers will be even more compelled to settle in Europe or the US, the taxes to support non-working parts of the population are leaving less margin for starting businesses and families. So american companies won't be outsourcing for much less for same quality, at least from latin america. I think that the US won't need to be too stringent on foreign remote workers: third-world politicians will push workers out in their thirst for taxing income (and the rationale will be: "They have to pay their fair share!").
Depends on the role and the sector in tech you work in. IMO, most HN commentators seem to be generic frontend/backend/fullstack devs working on generic CRUD apps at a handful of FAANGs. With the right domain knowledge, you could earn way more in other sectors or niches within tech even with those skills themself.
Do you have any examples of the type of niches those skills would transfer to?
so, ceo's are trying to end remote tech workers domestically, but for less cost, are open to workers even farther away.
Pardon me, but after 13 years in business I have to ask, why are we calling cheaper workers "talent" ?
Because when corporations want to work in unison to scare the workforce into believing they’re worth less so you can pay them less, it’s important to use words that build the brand you’re going for.
For example, if you said “companies are looking to countries without the mental workforce to do their jobs in a bid to save money”, your argument would be weak. People would see your propaganda, laugh, and then look for a new job at another company to jump ship and get a big pay increase.
If instead you refer to those people as “talent”, there’s a small but non negligible percentage of the population who is sheepish enough to become afraid and form some kind of stockholm syndrome to endear them to you as their employer.
For a similar exercise, note how onlyfans doesn’t refer to its workers as “sex workers”, “whores”, or “prostitutes”, but rather as “content creators”.
As long as they can do the job, they are talent. What it costs to hire them is an orthogonal issue.
The same way we call SWEs in Cary, NC talent even though they earn half of what they would for equivalent roles in the Bay Area (compare RTP salaries for the exact same roles and levels at Cisco, Netapp, etc). The value of labor is always an arbitrage between talent, location, supply, and demand.
Don't count on it.
The premise is very incorrect. Latin America has a dearth of educated workers. It's why their societies' infrastructure is barely functional.
I've spent 1.5 years living in Mexico. During that time, I met one electrical engineer, one software engineer, and one digital marketer (Mexican citizens). I am sure I would meet more if I spent more time in Guadalajara or Mexico City. But, I wouldn't expect to replace a labor force in any useful amount.
95% of people I met didn't seem to be very skilled at reading or writing, as though they barely had middle-school level education. I expect the same is true for most Latin American countries except Chile & Costa Rica, which have higher rates of human development.
Notice there is not much data in this article. It's just a qualitative description of one person's experiences in hiring. And discusses a survey of 18 companies trying to hire in Latin America.
The main export of Costa Rica now are services, several of the largest tech companies in the world are in Costa Rica (Intel, Amazon, Microsoft, HP, Dell, IBM, Oracle, etc...) and there are hundreds of software and web development companies, so when the article talks about it's hard to find talent in Latin America, I don't know WTH it's talking about.
Wow. You're right. I didn't believe services are Costa Rica's largest export, had to look it up.
In retrospect, it makes sense: English is widely spoken in Costa Rica. And along with Chile, it has the highest rate of human development in Latin America.
"In 2019, Costa Rica exported $9.38B worth of services."
"The top exports of Costa Rica are Medical Instruments ($3.27B), Bananas ($1.28B), Tropical Fruits ($1.03B), Integrated Circuits ($1.02B), and Orthopedic Appliances ($850M),"
I think VMWare played a massive role in building the talent scene in Costa Rica.
I've heard down the grapevine that an average dev's salary in Costa Rica would be significantly higher than in MX, though CDMX and Monterrey supposedly trounces San José as a city.
I’ve never been to Mexico, but on the topic of some other countries in Latin America I believe this to be widely inaccurate.
Also going like “lack of educated people is why their society doesn’t function” is pretty crazy. Institutions are the reasons societies work or not.
Let's follow your logic.
Let's assume Well Functioning institutions are in place.
Let's take it one step back, and one step forward from there:
- one step back: What enables Functioning Institutions? The institutions are composed of people. Good institutions require educated, skilled professionals.
- one step forward: Once we have Well functioning institutions... Well, what do they exist for? To support a functioning society, perhaps? Which is dependent on what? (Institutions, yes, we just covered that-- What next? .....oh that's right.... educated, skilled professionals)
To say "institutions are the reason <thing> functions" is quite a myopic perspective. Did an institution install the plumbing in your house? Did an institution grow the food you eat? :) I only ask that you step through your own logic.
Uhhh - I think you're overgeneralizing here in a big way.